Procyclicality of the Comovement between Dividend Growth and Consumption Growth
Journal of Financial Economics, forthcoming
Duffee (2005) shows that the amount of consumption risk (i.e., the conditional covariance between market returns and consumption growth) is procyclical. In light of this "Duffee Puzzle", I empirically demonstrate that the conditional covariance between dividend growth (i.e., the immediate cash flow part of market returns) and consumption growth is (1) procyclical and (2) a consistent source of procyclicality in the puzzle. Moreover, I solve an external habit formation model that incorporates realistic joint dynamics of dividend growth and consumption growth. The procyclical dividend-consumption comovement entails two new procyclical terms in the amount of consumption risk via cash flow and valuation channels, respectively. These two procyclical terms play an important role in generating a realistic magnitude of consumption risk. In contrast to extant habit formation models, the conditional equity premium no longer increases monotonically when a negative consumption shock arrives because it might lower the amount of risk while increasing the price of risk.
◘◘ 2018 E(astern)FA (2018/04), 2018 MFA (2018/03), 2017 SoFiE Conference (main conference), New York (2017/06), Federal Reserve Bank of New York, New York (2017/06), 2017 AEA/AFA/ASSA (poster presentation), Chicago (2017/01), 28th Australasian Finance and Banking Conference (AFBC), Ph.D. Forum (2015/12), 28th AFBC, Asset Pricing II (2015/12), Ph.D. Seminar, Columbia Business School (2015/11), 15th Transatlantic Doctoral Conference, London Business School (2015/05), Third-year paper presentation, Columbia Business School (2015/01)
◘◘◘ winner of 28th AFBC 2nd best paper at the Ph.D. Forum (2015/12)
 Global Risk Aversion and International Return Comovements
I establish three stylized facts about global equity and bond return comovements: Equity return correlations are higher, asymmetric, and countercyclical, whereas bond return correlations are lower, symmetric, and weakly procyclical. To interpret these stylized facts, I formulate a dynamic no-arbitrage asset pricing model that consistently prices international equities and bonds; the model features various time-varying global macroeconomic uncertainties and risk aversion of a global investor. I find that different sensitivities of equity returns (strongly negative) and bond returns (weakly positive or negative) to the global risk aversion shock can explain the observed comovement differences. Global risk aversion explains 90% (40%) of the fitted global equity (bond) comovement dynamics.
◘◘ Annual Workshop on Investment- and Production-Based Asset Pricing, Olso (scheduled), 2020 AEA (2020/01), Stanford SITE "Session 8: The Macroeconomics of Uncertainty and Volatility" (2019/08), 2019 UBC summer conference (2019/07), University of Zurich (2018/12), University of Luxembourg (2018/12), London Business School (2018/09), 2018 E(uropean)FA (2018/08), 2018 CICF (2018/07), Boston College (Carroll), Cornerstone, Emory (Goizueta), Georgetown (McDonough), Goldman Sachs, Johns Hopkins University (Carey), University of California (Riverside), University of Minnesota (Carlson), University of Notre Dame (Mendoza), University of Oklahoma (Price), University of Southern California (Marshall), University of Wisconsin Madison, Finance Ph.D. Seminar, NYU Stern (2017/12), Finance Faculty Free Lunch, Columbia Business School (2017/11), Ph.D. Seminar, Columbia Business School (2017/10), Financial Economics Colloquium, Econometrics Colloquium, Columbia University (2017/10, 11), Federal Reserve Bank of New York, New York (2017/09)
◘◘◘ Dissertation Award, Federal Reserve Bank of New York, New York 2017
 The Time Variation in Risk Appetite and Uncertainty
Management Science, submitted
We formulate a dynamic no-arbitrage asset pricing model for equities and corporate bonds, featuring time variation in both risk aversion and economic uncertainty. The joint dynamics among cash flows, macroeconomic fundamentals and risk aversion accommodate both heteroskedasticity and non-Gaussianity. The model delivers measures of risk aversion and uncertainty at the daily frequency. We verify that equity variance risk premiums are very informative about risk aversion, whereas credit spreads and corporate bond volatility are highly correlated with economic uncertainty. Our model-implied risk premiums outperform standard instruments for predicting asset excess returns. Risk aversion is substantially correlated with consumer confidence measures, and in early 2020 reacted more strongly to new Covid cases than did an uncertainty proxy.
◘◘ Chicago Fed seminar (2020/06), 8th HEC-McGill Winter Finance Workshop (2020/03), 9th ITAM Finance Conference (2020/02), 2019 EFA (2019/08), 2019 CICF (2019/07), 2019 EFMA (2019/06), 2019 FIRS (2019/05), 15th European Winter Finance Summit (2019/03), 2019 MFA (2019/03), 2019 AFA (2019/01), 31st Australasian Finance and Banking Conference, Sydney (2018/12), 2018 CIRF (2018/12), University of Zurich (2018/12), University of Luxembourg (2018/12), 2018 NFA (2018/09), "Machine Learning and Finance: The New Empirical Asset Pricing" hosted by University of Chicago Booth (2018/07), 2018 North American Summer Meeting of the Econometric Society (2018/06), 11th Annual SoFiE Conference (2018/06), Baruch College (2018/05), Federal Reserve Board's Conference on Risk, Uncertainty, and Volatility (2018/04), Columbia Women in Economics (2018/03), Columbia Business School (2018/03)
◘◘◘◘ winner of Global association of research professionals (GARP) research excellence award, China International Risk Forum (2018/12)
 Variance Risk Premium Components and International Stock Return Predictability
We decompose the U.S. variance risk premium (VP) into its downside and upside components (DVP and UVP, respectively) as proxies for asymmetric global risk variables, and find that acknowledging for asymmetry in VP significantly improves its international stock return predictability. To rationalize our empirical findings, we propose an international dynamic asset pricing model featuring asymmetric non-Gaussian shocks and partial global integration. We find that (i) DVP (UVP) is mostly driven by global risk aversion (economic uncertainty), (ii) international equity risk premiums exhibit distinct loadings on global premium determinants, and (iii) DVP (UVP) transmits to international markets through financial (economic) integration.
◘◘ 2021 AFA (scheduled), IFABS 2019 Medellín Conference (2019/12), Stanford SITE "Session 7: Asset Pricing Theory" (2019/08), NASMES Summer Meeting (2019/06), ECWFC@WFA (2019/06), FMA Global Conference in Latin America (2019/05), E(astern)FA 2019 (2019/05), MFA 2019 (2019/03), Federal Reserve Board (2019/03), Econometric Society European winter meeting 2018 (2018/12), 2018 CIRF (2018/12), Boston Macro Juniors Workshop (2018/11), Boston College Carroll (2018/11)
◘◘◘ Semifinalist, 2019 FMA Global Conference Best Paper Awards
 Risk, Uncertainty and Monetary Policy in a Global World
Since the global financial crisis, there has been renewed interest in understanding how monetary policy shocks transmit across countries through risk variables, spurring a literature on the ``global financial cycle.'' This paper studies how (conventional and unconventional) monetary policy shocks affect risk and uncertainty in three large economies: the US, euro area, and Japan. We construct measures of financial risk factors for each country by decomposing option-implied variances of the stock market into a conditional variance component (``uncertainty'') and a variance risk premium component (which is more closely associated with risk aversion). The conditional variance component is computed using a novel non-linear forecasting model. We then analyze monetary policy effects on risk and uncertainty. We do not find evidence that US monetary policy affects risk or uncertainty in the other countries. Instead, we provide evidence of significant spillovers through interest rates. We find limited evidence of cross-country monetary policy spillovers to the major asset classes (stocks, bonds, exchange rates, commodities), but rather strong evidence of non-monetary policy-driven risk aversion and uncertainty shocks transmitting across all countries (not just emanating from the US).
◘ Paper [New]
◘◘ 2021 AEA (scheduled), 2020 Annual Meeting of the Central Bank Research Association (2020/09), BI Olso (2020/08), University of Cincinnati (2020/03), 11th International Research Forum on Monetary Policy (IRFMP) (2020/03), MFA (2020/08), SNB-FRB-BIS High-Level Conference on Global Risk, Uncertainty, and Volatility (2019/11), 20th IWH-CIREQ-GW Macroeconometric Workshop: Micro Data and Macro Questions (2019/10), Conference on Advances in Applied Macro-Finance, Istanbul, Turkey (2018/12)
 Risk Aversion Propagation: Evidence from Financial Markets and Controlled Experiments
(with Xing Huang)
While the time variation in risk appetite is widely examined, there is scant research on how risk appetite may respond in an international context. We study risk aversion (RA) propagation from US to other major developed economies using both empirical data and controlled experiments. First, by exploiting daily financial market and news data between 2000 and 2017, we identify US risk aversion events -- both high and low -- and show that the international pass-through of US high RA events is significantly higher (62%) than that of US low RA events (43%), suggesting asymmetric US risk aversion propagation. Next, in three controlled experiments, we find evidence that non-US subjects when primed with a US financial bust shock exhibited asymmetrically higher negative emotions and risk aversion, suggesting that an "emotion"-related non-fundamental channel may play an important role in the excessive US high RA propagation.
◘ Paper [Soon]
◘◘ SAIF (scheduled), Fudan University Economics (scheduled), SMU (scheduled), Boston College Brownbag (scheduled), Washington University in St. Louis Brownbag (2020/10), WAPFIN @ Stern, New York (2018/09)
◘◘◘ winner of Boston College Research Expense Grant, 2018-2019
 Uncertainty Shocks and Personal Investment: Evidence From a Global Brokerage
We use novel data from a global retail brokerage to study how shocks to uncertainty affect personal investment around the world. We consider three empirical uncertainty shocks that have been proposed by the literature — terrorism, natural disasters, and large stock price jumps. We then consider how within-country uncertainty shocks affect investment and delegation in global assets on the brokerage. Importantly, the within-country uncertainty shocks (e.g., a natural disaster in Spain) are unlikely to affect world asset fundamentals (e.g., the SPX500). This allows us to isolate the effects of uncertainty shocks on personal risk aversion from their effects on asset fundamentals. We find that uncertainty shocks have scant effects on personal investing. They primarily affect delegation to asset managers on the brokerage, but that the direction of the effects depends on the type of uncertainty shock. Investors increase their delegation by 5% following terrorist activity and reduce delegation by 8% following positive and negative stock market jumps. These findings suggest that the exogenous uncertainty shocks proposed by the literature have heterogeneous effects on individual investment.
◘◘ 2019 ANU-RSFAS Research Camp (2019/12), Boston College Brown Bag (2019/11)
◘◘◘ INQUIRE Europe Research Grant, 2020
 Bond Home Bias
The three stylized facts as established in Xu (2017b) suggest that international bond investment for a global (U.S.) investor is potentially more attractive from the perspective of diversification benefits, which at first glance seems to deepen another little-known puzzle: bond home bias is significantly higher than equity home bias (Coeurdacier and Rey, 2013).
 Investor Attention and Equity Risk Premium
(with Melk Bucher)
Traditional asset pricing models assume that information is instantaneously incorporated into prices when it arrives, which immediately assumes investor full attention during all times. Da, Engelberg, and Gao (2011) however show direct evidence that investors appear to have limited attention and it is time-varying. Our paper formally proposes and estimates a measure of investor attention arising from a conditional asset pricing model using a large cross-section of stocks. Our measure allows an endogenous attention to reflect the empirical fact that investors choose its appropriate degree in the wake of changing fundamental (state) variables, such as e.g. information flow.
 Growth Dynamics at Different Stages of Development
(with Geert Bekaert)
This paper characterizes growth rates at different stages of economic and financial development of 180 countries over 55 years (1960-2014). We present several static stylized facts. In particular, low development stage appears with high growth volatility and positive growth skewness, which is potentially caused by significant growth spurts in these country-years.